Sound Target
Increasing ticket sales with online ads delivered to real customers
About product
Music taste-based online ads
A lot of people listen to music online every day. There are many things we can tell about the person listening to particular artists and genres. Why don't we use it in online ads then?
Sound Target makes it possible. It's that simple - find your audience, parse their music and segment them for specific targeting in social media.
The most obvious customer for these kinds of ads is the b2b segment in the music industry. We've started with couple of big concerts and quickly got the results - at least 1% increase in CTR and further increase in CR on the website. Online tickets sellers were happy with our solutions, but the market in Russia and CIS is too small. The reason is that in 2016 there was only one social network with audio in user profiles - VK.com
Actually, it was not possible to get enough data through the VK API to analyse it, but since most of the user profiles were open for everyone, it was not too complicated to parse them. You just need a good web developer and enough money on your AWS account.
The thing is - you cannot do the same with all the services that could give you much more data - YouTube, Sound Cloud, Apple Music, Spotify and etc. So you can actually forget about such a business model anywhere else.
What would you do in this case? We decided to use the same data source we had before but try to find a different type of customer, which we did successfully.
The next test we did to try to analyse the audiences of Russian retailers in order to suggest them relevant music that should be played in shops. After couple of pilot launches with some retail shops we quickly realised that there's a huge lack of data for this kind of analysis - most of small shops simply don't have their communities set up on VK so you always have to make a lot of guesses about their real audience, which ruins the final results. Moreover, even if the shop has big community of its buyers, it is much smaller than an artist's audience, so it's really hard to get some relevant results.
Product timeline
How it all began in Jul 2016
One famous concert agency told us that they really needed to extract the audience who listens to the artist they were managing upcoming concert for. The statement was: "No one in the market can do it for us" so we decided to do it, which was completely successful.
Some stable growth in Aug - Sep 2016
While getting more and more customers, we understood that was better to do the same thing in foreign markets, but quickly realised it wasn't possible due to tech reasons. Time to search for new customers in Russia
Attempts to conquer retail in Oct - Nov 2016
Maybe if a person listens to his/her favourite music in the shop, it might increase his motivation to buy more. Well, it could be possible, but the lack of data we could collect for retail visitors ruined this hypothesis completely
Sudden competition and time to close - Dec 2016
While still having some big customers from music industry, for some reason we forgot to keep an eye on our competitors (simply because we got used to the fact that we had none). Eventually, a strong one arose who was doing the same thing with an extra 99 additional useful features, 10x cheaper for monthly subscription than to buy only 1 analysed user database from us.
The market usually switches to such alternatives quit fast and our situation just proves this fact. We decide to close this project because there was no reason to put in that much effort to compete in a very small market.
Some results
Customers
8 promoters
4 artists
2 venues
1 music festival
Efficiency
1% CTR increase in average
LTV
Up to 5 purchases per customer
CAC
Less than $10 in online ads
Revenue
$2700 in total
Costs
Almost $0. Team was driven by enthusiasm, platform operating on $10,000 grant from AWS.
Personal data & privacy
One last thing I should definitely mention. We've never misused any personal data; all the information used was publicly available for each Internet user. We've never used any private data like name, e-mail or phone number because we never had an access to it. The only thing being used in this case is social network user id that goes directly to advertising campaign audience. All of these actions are within social media privacy policies.
We do not support personal data misuse and we always act within the law.